November 2005 - Boost-users - lists.preview.boost.org

multi_array - changing number of elements
by fritz.birnbacher＠t-online.de 21 Nov '05

21 Nov '05

Hello, how can I change the number of elements in an array ? I want to do following: typedef boost::multi_array<double, 2> array_type; array_type::extent_gen extents; array_type A1(extents[2][3]); array_type A2; A2 = A1 best regards Fritz Birnbacher

2 2

[BGL, Graph] choose the correct container to handle graph with a property
by Johan Oudinet 21 Nov '05

21 Nov '05

Hello, I have a graph representing the Network topology and I want to replace it by a BGL graph. My problem is that I need to find a vertex by its IP address (an internal property), but I think we can only find a vertex with its vertex_descriptor. A solution is to maintain a map from IP address to vertex_descriptor. But removing a vertex with vecS for VertexList invalidates all vertex descriptors. Unfortunately, I need some efficiently to find a vertex, so listS seems not appropriated. Finally, I try to use setS for VertexList (and it works IMHO), but in the documentation: "The container must model Sequence or RandomAccessContainer." Hence, my questions are: - Can I use setS for VertexList ? And if I can't, why ? - Is there a better solution to find vertex by an internal property that handle an exterior map ? Cheers, -- Johan

2 2

Re: [Boost-users] Unknown program_options bomb
by Merrill Cornish 21 Nov '05

21 Nov '05

Volodya, Sorry about the formatting. It annoys me too. However, I'm using Earthlink's online Web Mail, so I'm stuck with some of their choices. There new mail (available RSN) may have a formatting option. RE: exception DOH! I should have thought of that. For everything else, I'm always very careful to check for exceptions. In this case, I tacitly assume it worked like getopts and just left unknown options on the command line. (Old habits die hard.) I haven't got time to try now, but I will put in the exception handlers and see what it's complaining about with my positional parameter. I realize that "bomb" wasn't very precise, but it was hard to describe what happened. First, Zone Alert would become excited and claim that my program was attempting to do something "dangerous". (Yes, that should have clued me in to an exception being thrown, but I was dense.) If I told Zone Alert to let it happen, then Windows popped up a dialog offering to send a report to Microsoft about my incompetence. :-) Thanks for your help. Merrill

1 0

boost::thread - mistake in documentation
by Oliver.Kowalke＠infineon.com 21 Nov '05

21 Nov '05

Hello, the documentation (http://www.boost.org/doc/html/condition.html) of class condition contains a mistake: instead of template<typename ScopedLock, typename Pred> bool timed_wait <http://www.boost.org/doc/html/condition.html> (ScopedLock&, Pred); it should be template<typename ScopedLock, typename Pred> bool timed_wait <http://www.boost.org/doc/html/condition.html> (ScopedLock&, const boost::xtime <http://www.boost.org/doc/html/xtime.html> & , Pred); so long, Oliver

1 0

contribution - sync. queue between producer/consumer threads
by Oliver.Kowalke＠infineon.com 21 Nov '05

21 Nov '05

Hello, I would contribute this code. The queue synchronizes the access between producer threads (putting items into the queue) and consumer threads (removing items from queue). I would be glad if you add this code to boost::thread library. With regards, Oliver <<queue.hpp>> <<test.cpp>>

1 0

Re: [Boost-users] Unknown program_options bomb
by Merrill Cornish 21 Nov '05

21 Nov '05

Will, I've changed my original po::store(parse_command_line(argc, argv, desc), vm); to po::store(po::command_line_parser(argc, argv). options(desc). positional(pd). run(), vm); where desc is is an options_description and pd is a positional_options_description. This looks like an exact match for po::variables_map vm; po::store(po::command_line_parser(ac, av). options(desc).positional(p).run(), vm); po::notify(vm); which is the example from http://www.boost.org/doc/html/program_options/tutorial.html. However, the code still bombs if I give it a positional parameter or an option it isn't expecting. Merrill

2 1

Re: [Boost-users] Unknown program_options bomb
by Merrill Cornish 21 Nov '05

21 Nov '05

Liam, I think you're right, but I haven't been able to figure out the exact incantation. :-) Merrill

1 0

Unknown program_options bomb
by Merrill Cornish 21 Nov '05

21 Nov '05

I am using Boost 1.33.0 with MinGW on Windows XP. I am using the following program_optins code to collect command line arguments: po::options_description desc("Options"); desc.add_options() ...[everyting defined in here works as advertised]... ;//add-options() // specify the single, required graph file positional parameter po::positional_options_description pd; pd.add("graph_file", 1); po::variables_map vm; po::store(parse_command_line(argc, argv, desc), vm); po::notify(vm); ... The above code is shamelessly lifted from the tutorial. Everything defined in desc.add_options() works including the "-h" alternative to "--help". However, two things fail. * vm.count("graph_file") returns 0 regardless of whether there are positional argument(s). * specifying an -x, --xx, --xx=n option where "x" or "xx" is not included in desc.add_options() causes store() to not return (the program simply stops). Am I missing something? Merrill

3 2

[wave] question about preprocessing hooks
by Alexey Bakhvalov 20 Nov '05

20 Nov '05

Hello, What is a correct way to implement 'defined_macro' hook? For some reason I always get empty 'definition' list. My hook implementation looks like this: class hook : public boost::wave::context_policies::default_preprocessing_hooks { public: template <typename TokenT, typename ParametersT, typename DefinitionT> void defined_macro(TokenT const &macro_name, bool is_functionlike, ParametersT const &parameters, DefinitionT const &definition, bool is_predefined) { if (!definition.empty()) { // never gets here } } }; I am using Boost 1.33 and VC .NET 2005 compiler. Alexey Bakhvalov

2 1

Re: [Boost-users] Wave library fails to compile with VC7.1 - Bufferoverrun in cl.exe
by Hartmut Kaiser 20 Nov '05

20 Nov '05

> Didier Malenfant wrote: > > > It seems like this was fixed by turning off the rtti for > > VC7.1 and I verified that both V1 and V2 jamfiles have the > workaround > > line in them (not sure which one is getting used as I'm not passing > > any parameters for that) > > V1 is used by default. Just another thought. Since you're using the 'vc-7_1-stlport' toolset, you'll need to add a <vc-7_1-stlport><*><rtti>off # workaround for compiler bug line to the Jamfiles. Could you please try this and report back? Regards Hartmut

1 0